[ADMB Users] crappy software

[dave fournier]

OK I have 280 simulations which is enough to
see the pattern.  lmer has a 23% type 1 error
while admb-re has a 7.5% type 1 error.

I wonder how many papers have used this
method?   It makes Excel look good!




                     % type 1    num sims   mea n     std dev        
mean est std
                     error                                b1            
b1           in the model

lmer            0.232143       280           0.987887 0.190596    0.118172

ADMB-RE    0.075              280          1.00791    0.187442    0.162416